New research has called into question the assumption that spending more on better IT will improve productivity and lead to higher profits.
For years politicians and technology companies have pointed to the boom experienced in the US in the late 1990s as evidence that more IT means more productivity.
From 1995 onwards, spending on IT by US companies has doubled. By the end of 1999 US companies were shelling out an average of $3,000 (£2,085) on IT for each employee.
At the same time, the US enjoyed some of the most spectacular economic growth it has seen so far. Growth in labour market productivity rose from 1.4% to an unprecedented 2.5% a year.
But research has cast doubt on the central role played by IT in this economic transformation. A study from McKinsey Global Institute, research arm of the management consultancy firm, suggests that the bulk of the improvements were due to changes in the way companies delivered their products and services.
IT, though important, was not the main driving force.
Although McKinsey based its conclusions on an analysis of US government data, its findings are just as relevant to IT directors in the UK.
The group found that nearly all the post-1995 growth in the US economy came down to just six sectors of the economy - retail, wholesale, securities, telecoms, semiconductors and computer manufacturing. The remaining sectors of the economy showed virtually no productivity improvements, or even a decline. And yet they accounted for 62% of the growth in IT spending.
On closer inspection, McKinsey found a host of reasons for the boom in the six key areas of the economy, many of them only remotely linked to IT.
In the semiconductor industry, productivity jumped from 43% to 66% largely because of improvements in chip manufacturing processes. Intel, driven by competition from rival AMD, shortened the time between new chip versions and began improving the performance of each new chip more rapidly.
In telecoms, the Government's decision to license a new spectrum for mobile phones increased competition. This led to lower prices, more customers, and allowed the industry to spread its costs over a wider customer base.
Even in retail, IT played only a secondary role. Here, the driving force was the US budget supermarket Wal-Mart. The retailer saw its market share grow from 9% in 1987 to 27% by the mid-1990s, driven by productivity figures that were 40% higher than the industry average.
Competitors took up many of Wal-Mart's innovations. They included the use of electronic data interchange and wireless barcode scanning, but some of the most significant changes were not IT related, such as economies of scale in the stores and warehouse logistics.
In other areas, investments in IT have yet to show any payback. Hotels have spent money on reservation systems, but they have led to only marginal improvements. Investments by retail banks in Internet banking have yet to reap dividends. And long-distance telecoms companies have invested in networks that are likely to remain underused for years to come.
The existence of these sectors, according to McKinsey, shows that IT alone is not a magic solution that can drive productivity growth.
It is only when IT enables significant management innovations that it plays a major role in driving productivity.
"IT directors will need to realise that IT is not a silver bullet and that by spending on IT alone they will not get improvements. It is only when you have IT investment and managerial innovation that changes the way a business is run that you will increase productivity," said James Manyika, a partner at McKinsey.
In fact, many IT investments in the late 1990s were made to maintain existing capabilities, rather than to generate rapid returns. Companies spent millions upgrading systems for Y2K. They invested in Internet and network technologies with long-term returns in mind, and upgraded PCs to keep up with changing standards, rather than to improve efficiency.
McKinsey's findings were backed up by research by Butler Group earlier this year. Butler analysed the IT investment of more than 1,500 UK companies and concluded there is only a random link between how much a company invests in IT per employee and how much money it makes back through returns on shareholder equity.
Like McKinsey, Butler's research highlighted the need for companies to be well managed if they are to get a return on investment.
McKinsey's findings, however, do provide some valuable lessons for IT directors, said Manyika.
For instance, IT directors should make sure new systems are tailored to the way their company works. Those firms that benefited most from IT invested in technology that fitted with their existing business practices. Those that did not often had to change their business practices to match the IT.
"In a lot of sectors investments were made because there was a bit of an arms race going on. We need to buy this technology because the competition has bought it, but we are not sure how we are going to use it."
Whether IT spending patterns will return to anything like pre-2001 levels is difficult to predict, said McKinsey. It will depend how quickly economic confidence is restored and how long it takes before compelling new applications emerge on the market.
For copies of the report go to www.mckinsey.com